This paper presents a secure computation model for k-nearest neighbor (kNN) queries on encrypted databases. The authors propose a model called SCONEDB, which ensures secure computation on encrypted data. The main challenge is to perform kNN queries on encrypted data without revealing the original data. To achieve this, they develop an asymmetric scalar-product-preserving encryption (ASPE) scheme that preserves scalar products but not distances, making it resistant to level-2 attacks. They also extend this scheme to resist level-3 attacks with additional overhead. The ASPE scheme is designed to preserve type-2 scalar products (between database points and query points) but not type-1 or type-3 scalar products (between database points). This allows for distance comparisons without revealing the actual distances. The authors also introduce random splitting and artificial dimensions to further enhance security against stronger attacks. The proposed schemes are evaluated for performance and efficiency, showing that they can securely perform kNN queries on encrypted data. The results demonstrate that the schemes are resilient to various attack models, including level-2 and level-3 attacks.This paper presents a secure computation model for k-nearest neighbor (kNN) queries on encrypted databases. The authors propose a model called SCONEDB, which ensures secure computation on encrypted data. The main challenge is to perform kNN queries on encrypted data without revealing the original data. To achieve this, they develop an asymmetric scalar-product-preserving encryption (ASPE) scheme that preserves scalar products but not distances, making it resistant to level-2 attacks. They also extend this scheme to resist level-3 attacks with additional overhead. The ASPE scheme is designed to preserve type-2 scalar products (between database points and query points) but not type-1 or type-3 scalar products (between database points). This allows for distance comparisons without revealing the actual distances. The authors also introduce random splitting and artificial dimensions to further enhance security against stronger attacks. The proposed schemes are evaluated for performance and efficiency, showing that they can securely perform kNN queries on encrypted data. The results demonstrate that the schemes are resilient to various attack models, including level-2 and level-3 attacks.